Neural Network-based Fault Classification in Multi-Level Inverters

Multi-level inverters are widely used in power systems, power electronics and renewable energy applications. High power output, low value of total harmonic distortion THD are the major advantages of Multi-level inverters, which makes them compatible with variable speed applications. As the levels increase, there are more number of faults occurring in the switches. Early detection of faults are necessary to protect the equipment and prevent damage to the power electronic system. This paper focuses on neural network-based fault analysis of multilevel inverters. Comparative analysis of the voltage output under different fault conditions is simulated and results are elaborately presented.